Measurement of Multilayer Coating Thickness on Interwoven Carbon-Fiber-Reinforced Polymers by Sparse Deconvolution of THz Polarimetry Pulses

Terahertz time-domain spectroscopy (TDS) has shown promising applications in non-destructive thickness measurement of various industrial products. Despite significant advances, multilayered, optically thin, and anisotropic samples pose extra complications to the analysis of thickness measurements. Interwoven carbon-fiber reinforced polymers, which show polarimetric anisotropy and surface non-uniformity due to the weave structure of the carbon bundles, are among these complicated specimens. In this study, we use a dual-channel polarimetric THz-TDS scanner to address this problem for anisotropic structures. Use of the polarization information allows for mapping the thickness variation over the warp and weft of the samples. In addition, by utilizing the sparse deconvolution algorithm, we accurately predict the thickness of optically thin multilayer coating with an uncertainty (resolution) of ±12 μm and accuracy of 92%, which represents a significant improvement over the previously demonstrated performance metrics.